J Sci Food Agric 1997, 73, 301È306

Relationship between Hunter Color Values andb-Carotene Contents in White-Fleshed AfricanSweetpotatoes (Ipomoea batatas Lam)Michael A Amenya* and Paul W Wilsonb

a Department of Food Science,b Department of Horticulture, Louisiana State University Baton Rouge, Agricultural Center, LA 70803,USA

(Received 13 April 1995 ; revised version received 24 October 1995 ; accepted 29 August 1995)

Abstract : White-Ñeshed sweetpotatoes (Ipomoea batatas (L) Lam) are a majorfood crop in Uganda. This study was done to evaluate carotenoid content insweetpotatoes and to relate the color of the sweetpotatoes to the b-carotenepresent. A Hunter Color Tristimulus Meter was used to determine the color andthe b-carotene was determined by reverse-phase HPLC. Regression analysis wascarried out on the color values L, a, b, b/a and tan~1 b/a to determine whichcolor value could be used to express the content of b-carotene in white-Ñeshedsweetpotatoes. The “bÏ color value appeared to be the best estimation for corre-lation at 0É74 in this study for raw roots and 0É09 for pure� e. The “aÏ value was0É39 for raw roots and 0É16 for pure� e, L was [0É74 for raw roots and [0É45 forpure� e.

Key words : Hunter meter, carotenoids, b-carotene, sweetpotato, Uganda.

INTRODUCTION

Pre-formed vitamin A found in meat, liver and eggs isfrequently too expensive for the economically deprivedin developing countries (Ong and Tee 1992). Fortu-nately, many of these countries have abundant vegeta-tion rich in carotenoids. There should therefore be novitamin A deÐciency problems ; however, this is nottrue. Low intake of carotenoids, and diets lacking in fatand protein, make absorption of the precursors ofvitamin A very poor (Ong and Tee 1992). Roots andtubers are generally low in carotenoids. They are,however, consumed widely in many developing coun-tries and, therefore, their use as a source of carotenoidsmerits further studies (Ong and Tee 1992).

The sweetpotato (Ipomoea batatas) is a major foodcrop in developing and developed countries (Woolfe1992). Little dietary information is known about thewhite-Ñeshed varieties grown more commonly in

* To whom correspondence should be addressed at : Tech-nology, Makerere University, PO Box 7062, Kampala,Uganda.

developing countries (Woolfe 1992). Color intensity ofb-carotene in a food may be considered a reliable indi-cation of vitamin A value (Lauber et al 1967). Thus, it isappropriate to apply color measurements for rapid esti-mation of carotenoid content (Lauber et al 1967 ;McGuire 1992 ; Takahata et al 1993 ; Camire et al 1994).The advantages of this method are its rapidity and thecombination of quantitative and descriptive evaluationof the color in a single determination. It has been usedwith several food products (Kramer 1954 ; Worthington1954). Ahmed and Scott (1962) published correlationcoefficients showing a high degree of associationbetween Hunter color value “aÏ and carotenoid contentof raw roots and processed samples from a series ofbreeding lines of sweetpotatoes. Hunter color param-eters did not show any di†erences for transversal cuts inthe root (proximal, median and distal) (Almeida et al1992). Cooking, however, resulted in signiÐcantdecreases in the parameters, L (lightness) and ]a (red)and an increase of [a (green) parameters. The bparameter (yellow) decreased due to cooking in the caseof some lines.

301J Sci Food Agric 0022-5142/97/$09.00 1997 SCI. Printed in Great Britain(

302 M A Ameny, P W W ilson

Carotenoids in white-Ñeshed sweetpotato cultivarswere studied by Martin (1983) and Almeida and Pen-teado (1988). Although loosely called “white-ÑeshedÏ,these vary from white to cream, light yellow or lightorange. Cooking almost invariably intensiÐes the color.The cooked root may be white, grey, greenish, yellow-ish, yellow or orange. White-Ñeshed sweetpotatoes areusually preferred as a food in the tropics, because theydo not have a carrot-like sweet Ñavour (Martin 1983).The green color of the cooked sweetpotatoes may beassociated with an epoxide of the carotenoids (Ball1988).

The objective of this study was to determine therelationship between Hunter color reading and b-carotene content in raw and processed sweetpotatoes.The use of the Hunter Color Meter is being tested, sothat it may be used where HPLC is not available. Thefollowing studies were carried out to arrive at ananswer. Four cultivars of white-Ñeshed sweetpotatoes ofAfrican originÈT3013, T1702, T3002 and T3006Èwerechosen and carotene in them was evaluated.

MATERIALS

Sweetpotatoes (Ipomoea batatas Lam) were obtained astissue culture from the International Institute of Tropi-cal Agriculture (Nigeria) through the USDA-ARSSouthern Regional Plant Introduction Station. Allsweetpotato varieties were grown at Hill Farm(Louisiana State University, Baton Rouge, LA, USA),following common cultural practices. The sweetpotatoeswere harvested mechanically and the sweetpotatoes thatwere not damaged during harvest were chosen for thestudy. After harvest the sweetpotatoes were packed in18 kg crates and placed in a purpose-built curing roomfor 1 week. The curing room was at 30¡C and 80È90%relative humidity. This resulted in the healing of anywounds that were incurred during harvest and post-harvest handling. Curing provided a barrier to moistureloss and impeded microbial invasion of the tissue bymeans of suberisation. The sweetpotatoes were thenstored in a purpose-built storage facility at 15¡C and80È90% relative humidity for 1È6 months until neededfor the experiment. The four sweetpotato cultivarsranged in Ñesh color from white to yellow, and wereused for color and total carotenoid content determi-nation for both raw and pure� e samples.

METHODOLOGY

Processing into pure� e

Sweetpotato roots were sprayed with warm water toremove surface soil and lye peeled for 3È5 min in asodium hydroxide solution (150 g litre~1) at 80È90¡C.

They were rinsed in water to remove sodium hydroxideand coarsely chopped in a food cutter (Hobart Manu-facturing Co, Troy, OH, USA) to a 5È10 mm particlediameter. Fifteen percent of the chopped portion wasremoved to be added back later as a source of enzymefor the conversion of starch. The remaining choppedmaterial was cooked for 40 min in a stream-jacketedkettle at 90¡C and ground through a comminuting mill,with a 5 mm screen (Fitzpatrick Co, Chicago, IL, USA)to form a pure� e. The chopped raw portion was addedand the mixture held at 72È74¡C for 20 min, after whichthe temperature was raised to 90¡C to inactivateenzymes. The pure� e was milled using an 0É8 mm screen.Consistency was adjusted by adding water to achieve6È8 Bostwick at 90¡C. Cans (401] 411 mm) were Ðlledwith the pure� e and processed in a steam retort at 115¡Cfor 100 min. The canned pure� es were stored at 17È25¡C.

Carotenoid determination

Carotenoid content was determined as described byAmeny (1994). Sweetpotatoes were grated lengthwise ona cheese grater. Ten grams of each grated sample wasweighed in a homogenizer (Omni 17105, Omni Interna-tional Waterbury, CT, USA). To this was added 20 g ofanhydrous sodium sulphite, 1 g of magnesium carbon-ate and 100 ml of stabilised tetrahydrofuran (THF). Forthe processed pure� e, 10 g were weighed into the homog-eniser, followed by 20 g of anhydrous sodium sulphate,1 g magnesium carbonate and 100 ml of stabilisedTHF. The samples were homogenised for 5 min at aslow speed, then vacuum Ðltered through a Buchnerfunnel Ðtted with a Whatman Ðlter paper. The Ðl-d42trate was then brought to a Ðnal volume of 500 ml,evaporated to dryness at 40¡C over nitrogen in a rotaryevaporator (Buchi Laboratorimus-Technik AG, Flavil,Switzerland) and taken to a Ðnal volume of 10 ml withstabilised THF. The extracts were stored in brownbottles with TeÑon-lined caps under nitrogen in a coldroom at [20¡C until needed for injection into anHPLC, for spectrophotometric analysis and Huntercolor analysis. The term carotenoid as used refers to theTHF-soluble pigments exhibiting absorption at 450 nm.

Spectrophotometry

Spectrophotometric analyses were done on extractsusing a scanning UVÈVis spectrophotometer (PerkinElmer, Norwalk, CT, USA), to determine the maximumabsorption wavelength. Total carotenoids were thendetermined at the maximum wavelength of absorptionwith b-carotene as a standard. Stock solutions of b-carotene (Sigma Chemical Co, St Louis, MO, USA)were prepared by weighing 25 mg into a 100 ml brownlow actinic volumetric Ñask and bringing the volume to100 ml with THF. Four working standards were pre-

b-Carotene content of African sweetpotato 303

pared by taking 1, 2, 3 or 4 ml from the stock andmaking up to 100 ml. Standards were stored at [20¡Cunder nitrogen. A standard curve was prepared usingthe working solutions and the unknown b-carotene wasdetermined against the standard.

HPLC determination of b-carotene

A rapid non-aqueous reverse-phase HPLC was used,modiÐed from Bushway (1985). Conditions were asfollows : column packing, C-18 5k Vydac 201TP54 ;column, 150 mm] 4É6 mm; isocratic solvent,methanol/chloroform (90 : 10) ; Ñow rate, 1 ml min~1 at500 psi ; ambient temperature, 20¡C. b-Carotene wasused as standard to determine the retention time(Ameny 1994).

Hunter Color Laboratory measurements

Tristimulus instruments are designed to respond tospectral distributions of light in the same manner as thehuman eye. Duplication of the human eye response tolight is accomplished by adjusting the spectral responsesof photodetectors with Ðlters so that they are spectrallyequivalent to the International Commission on Illumi-nation (CIE) standard (Hunter and Harold 1987). A tri-stimulus instrument has, in addition to sourceÐlter-photodetector combinations for standard observersimulation, a device that computes directly chromaticdimensions of color, such as “aÏ, and “bÏ or “xÏ and “yÏ.Such an instrument with direct readings of “aÏ andRd ,“bÏ, or “LÏ, “aÏ and “bÏ scales is called a tristimulus color-imeter. The “LÏ measures from black (0) to white (100),]b measures yellow, [b measures blue, ]a is for red,while [a measures green colors.

Seven roots, approximately 300 g, of each sweetpo-tato cultivar were selected, washed, hand peeled with astainless-steel knife and cut transversely (discs) in themedian region for color measurements (raw roots). Thediscs were then boiled for 10 min and Hunter colorreadings were taken immediately.

Color was determined immediately on the freshly cuttransverse surface from the median portions of the rootusing a Hunter Color Laboratory Machine, ModelXL-23 Colorimeter (Gardner Instruments, GardnerLaboratory Inc, Bethesda, MD, USA), using WhiteStandard Tile XL-23-168-C Normal Beam, angle 45¡,opening 5 cm, and LL\ 92É94, aL \ [0É89 andbL\ 1É48. The outside (surface) color was measured byplacing the peeled central area of the unprocessed rooton the color meter opening. Extracts were obtainedfrom storage at [20¡C for color determination. Colorand carotenoid content were determined on cannedpure� e samples of four sweetpotato varieties after storagefor at least 1 month. The colors of pure� e samples wereevaluated as the average of duplicate samples from

seven cans from the same batch. Ten grams of the pure� ewas placed in a layer in the glass container and this wassufficient to allow di†erentiation of meter readings dueto the transmitted light. the hue angle was determinedas tan~1 b/a (Little 1975).

Statistical analyses

Regression analysis was carried out to determine therelationship between b-carotene and color values L, a,b, hue and b/a (SAS 1994). Analysis of variance of colorvalues was computed by GLM and the di†erencebetween means was determined by DuncanÏs test.

RESULTS AND DISCUSSION

The total carotenoids in the sweetpotatoes were thenanalysed for UVÈVis characteristics in comparison to ab-carotene standard at 450 nm to determine theapproximate total carotenoid content in the extracts ;the results are shown in Table 1. The total carotenoidcontents varied from 0 to 2600 kg kg~1 in the rawextracts and from 200 to 1900 kg kg~1 in the processedsamples. The b-carotene ranged from 14É3 to1260 kg kg~1, fresh weight basis (fwb). A comparisonwas carried out with work previously done on b-carotene and this is shown in Table 2.

Hunter Color Tristimulus meter readings and thecarotenoid content in white-Ñeshed sweetpotato

Analysis of the color values showed that there was asigniÐcant di†erence in the “LÏ (white) values (P\ 0É05)for the surface of hand-peeled and uncooked sweetpo-tatoes and the middle region. The “LÏ (white) values forthe middle region were higher than the surface values,implying that the surface is darker than the middleportion (Table 3).

The “aÏ (red) values of the surface and the middleregion were signiÐcantly di†erent (P\ 0É05) for all thecultivars. For cultivar T3002, the “aÏ (red) value for thesurface was higher than that for the middle region ;similar results were obtained for the “aÏ (red) values inT3013 and T3006. The “aÏ (red) value for T1702 for themiddle region was higher than the surface “aÏ (red) value,implying that the middle portion was redder than thesurface. Therefore, it probably has more carotenoids(Lauber et al 1967).

The “bÏ (yellow) values for the surface and the middleregion were not signiÐcantly di†erent for cultivarT3002, but were di†erent for cultivars T3013 andT3006, and similar for cultivar T1702. The color valuechanges may be due to changes in carotenoid concen-tration being di†erent in content in the middle andouter regions (Gross 1991).

304 M A Ameny, P W W ilson

TABLE 1Absorption wavelengths and total carotenoids in sweetpotatoesa

T otal carotenoid content (fwb) (kg kg~1)

Samples T reatment Range Mean Distinct Peak (nm)

T1702 Raw 2200È2600 2491É4 ^ 170É4a 428, 454Pure� e 900È1900 1571É4 ^ 681É8b 428

T3002 Raw 0È200 64É2 ^ 102É5f ÈPure� e 400È2500 915É7 ^ 808É2cb È

T3006 Raw 900È2400 1448É7 ^ 481É5b 428, 453Pure� e 700È1200 933É3 ^ 159É8c 428

T3013 Raw 300È500 431É4 ^ 103É0d ÈPure� e 200È500 245É7 ^ 100É9e È

a Numbers with the same following letters are not signiÐcantly di†erent at P\ 0É05.b Means for four readings, the rest are means for seven readings. DuncanÏs Multiple RangeTest.

Boiling for 10 min a†ected the color values for all thecultivars (Table 4). The “LÏ (white) values decreased onboiling, but were not signiÐcantly di†erent from thoseof the uncooked samples. The decrease could be due to

changes in carotenoids, caramelisation, oxidation orphenol action during boiling. The other cultivars ofsweetpotatoes still had signiÐcant di†erences betweenthe “LÏ (white) values despite the general decrease in “LÏ

TABLE 2Comparison of b-carotene contents of sweetpotatoes from various parts of

the world

Cultivar/Country b-Carotene (kg kg~1, fwb)

T1702 (raw, African) 104É0T102 (pure� e, African) 114É4T3006 (raw, African) 216É2T3006 (pure� e, African) 50É1Sweetpotato (fresh, orange, Australia) 66 000*Sweetpotato (orange, steamed, Indonesia) 1550*Sweetpotato (orange, fresh Indonesia) 7360*Sweetpotato (yellow, Vietnam) 21 900*

a From West and Poortuliet (1993).

TABLE 3External and internal color values of unprocessed cured sweetpotatoa

Samples L b/a T an~1 b/a a b

T1702 Middle 78É0cd 2É09 64É46 16É13a 33É76aSurface 76É6ef 3É68 74É83 9É4b 34É68a

T3002 Middlea 85É1a 13É1 85É8 2É01fg 27É43cdSurface 82É4ab 6É5 81É29 4É31g 28É15cd

T3006 Middle 78É31d 2É91 71É06 7É58cd 22É1eSurface 74É53b 3É81 75É31 8É75bc 33É44ab

T3013 Middle 81É12bc 7É79 82É69 3É55eg 27É68cdSurface 74É3 5É29 79É31 5É7dc 30É2b

a Figures in the same column with the same following letter are not signiÐcantlydi†erent at P\ 0É05.b Middle\ Inside portion, cut transversely. n \ 3.

b-Carotene content of African sweetpotato 305

TABLE 4Color values for boiled, pure� ed and extracts from pure� e and raw sweetpotatoesa

Samples T reatment n L b/a T an~1 b/a a b

T1702 Boiled 3 62É2h 5É11 78É93 6É2c 31É7abPure� e 7 31É9jk 1É81 61É17 8É2bc 14É9fExtractP 7 29É97jk 1É16 49É25 [1É43i [1É66jkExtractR 9 30É38jk [1É54 [57É0 [2É2i 3É7hi

T3002 Boiled 3 76É8ef 13É0 85É62 2É13hg 27É84cdPure� e 5 36É7i 2É86 70É7 4É36eg 12É48fExtractP* 7 32É6j 2É86 70É74 [1É2i [2É5jkExtractR* 7 31É58j 2É08 64É35 [1É71i [5É69jk

T3006 Boiled 3 63É61b 4É94 78É56 6É31dc 31É2bcPure� e 7 29É74k [6É1 [80É7 [2É16i 13É3fExtractP 3 30É9jk 0É84 40 [1É79i [1É51jkExtractR 11 31É5jk 0É16 9É4 [2É14i [0É36ij

T3013 Boiled 3 71É1g 8É11 82É9 3É5eg 28É40bcPure� e 7 33É7ji 2É21 65É71 5É59de 12É39fExtractP 7 30É2jk 3É96 75É8 [1É14i [4É52jkExtractR 11 32É64jk 4É77 78É18 [1É31i [5É4jk

a Figures with the same following letters in the same column are not signiÐcantly di†erent atP\ 0É05.b ExtractP* \ pure� e extract ; ExtractR\ Raw extract.

(white) value. Lye peeling and processing to pure� edecreased the “LÏ (white) value for all the samples,implying that the samples became less white or darkeron pure� eing. This was probably due to oxidation, phe-nolic action or caramelisation (Gross 1991). Theextracts both for pure� e and uncooked samples haddecreased “LÏ (white) values. There was no di†erence(P\ 0É05) in the “LÏ (white) values of the processed anduncooked extracts at P\ 0É05.

The “aÏ (red) values for T3002 and T3013 did notchange on boiling. In the other cultivars, T1702 andT3006, there was a decrease in the “aÏ (red) values. Allthe extracts had “[aÏ (green) values and there were nosigniÐcant di†erences among them at P\ 0É05. The “aÏ(red) value for the pure� e also decreased, as compared tothat of the unprocessed sweetpotatoes, that for T3006becoming “[ aÏ (green). Thus, the processed samplesbecame less red and more green probably due to isom-erisation of carotenoids, or complexing of carotenoidswith proteins (Gross 1991).

The “bÏ (yellow) value for T3002 was unchanged onboiling. It increased for T3013 and T3006, and forT1702 it decreased. All the extracts had “[bÏ (blue)values meaning the extracts were more blue, probablydue to complexing with proteins by the carotenoids(BauernÐend 1972). The pure� e values decreased forT3002, T3013, T1702 and T3006. Thus, pure� eing makesthe sweetpotato less yellow, probably due to cara-melisation or carotenoid loss (Gross 1991).

The hue angle of the middle region for cultivarsT3002 and T3013 was higher than that of the surfaceregion. For cultivars T1702 and T3006, the hue angle

for the middle region was less than that for the surfacearea. On boiling the hue angle for cultivar T3002 didnot change ; the value for T3013 decreased andincreased for T1702 and T3006. On pure� eing, the hueangles for T3002, T3013, T1702, T3006 all decreased.The hue angles for the extracts for T3002, T3013, T1702and T3006 were all lower than those of the correspond-ing raw & pure� e samples, but the decrease was greaterfor the raw extract than the processed extract in T3002,T1702 and T3006, while the processed extract decreasedmore than the raw extract in T3013. The change in hueangle may be due to loss, oxidation or isomerization ofcarotenoids, caramelisation or to enzyme action (Gross1991).

Relationship between color values L, a, b, b/a and tan~1b/a and b-carotene content

Regression analysis was carried out on the color valuesL, a, b, b/a and tan~1 b/a to b-carotene. Correlationanalysis was carried out to determine the strength of thelinear relationship between the two variables, color andb-carotene content. The results are shown in Table 5.The color value “bÏ (yellow) had the highest correlationcoefficient of 0É74 (P\ 0É05) showing that the relationbetween b-carotene and color value “bÏ (yellow) is linear,followed by “LÏ (white), with r \ [0É74. The value “aÏ(red) had r of 0É39, showing very little associationbetween color value “aÏ and carotene content. The “bÏ(yellow) color value appears to be the best measure forcorrelation between color value and b-carotene concen-tration. Takahata et al (1993) found the color value “aÏ

306 M A Ameny, P W W ilson

TABLE 5Correlation coefficients for color values L, a, b, b/a and tan~1b/a and b-carotene from the HPLC in raw and pure� e extract

n T reatment Color value r

11 Raw L [0É749 Pure� e L [0É45

11 Raw a 0É399 Pure� e a 0É16

11 Raw b 0É749 Pure� e b 0É09

11 Raw b/a 0É249 Pure� e b/a 0É29

11 Raw tan~1 b/a [0É389 Pure� e tan~1 b/a 0É19

(red) to be most appropriate for estimating b-carotenein orange and yellow Ñesh sweetpotatoes.

ACKNOWLEDGEMENTS

This work was supported by Louisiana MethodistChurch, Gerber Foods and Louisiana State University,Agricultural Center. This work was part of one of theauthorÏs (MAA) PhD dissertation.

REFERENCES

Ahmed E M, Scott L E 1962 A rapid objective method forestimation of carotenoid content in sweetpotato roots. ProcAm Soc Hort Sci 80 497È506.

Almeida L B, Penteado M V C 1988 Carotenoids and provita-min A value of white Ñeshed Brazilian sweetpotatoes(Ipomoea batatas Lam.). J Food Comp Anal 1 341È352.

Almeida L B, Penteado M V C, Ferreira V L P 1992 Relation-ship between carotenoid content and Hunter color param-eters of Brazilian sweetpotato (Ipomoea batatas Lam.).Revista Espanola de Ciencia y T ecnologia de Alimentos 32(6) 611È619.

Ameny M A 1994 E†ect of processing on protein and carot-enoids in weaning food formulated from sweetpotatoes orApios americana. PhD dissertation, LSU Baton Rouge,USA.

Ball G F M 1988 Fat Soluble V itamin Assays : A Com-prehensive Review. Elsevier, London, UK.

BauernÐend J C 1972 Carotenoid vitamin A precursors andanalogs in foods and feeds. J Agric Food Chem 20 456È473.

Bushway R J 1985 Separation of carotenoids in fruits andvegetables by high performance liquid chromatography. JAgric Food Chem 34 409È412.

Camire M E, Ismail S, Work T M, Bushway A A, HattermanW A 1994 Improvements in canned lowbush blueberryquality. J Food Sci 59 (2) 394È415.

Gross J 1991 Pigments in V egetables, Chlorophylls and Carot-enoids. Van Nostrand Reinhold, New York, USA.

Hunter R S, Harold R W 1987 T he Measurement of Appear-ance (2nd edn). John Wiley and Sons, New York, USA.

Kramer A 1954 Color Dimensions of Interest to the Con-sumer. Symposium of color in foods. Quartermaster Foodand Container Institute for the Armed Forces, USA.

Lauber J T, Taylor G A, Drinkwater W O 1967 The use oftristimulus colorimetry for the estimation of carotenoidcontent of raw sweetpotato roots. Proc Am Soc Hort Sci 91472È477.

Little A C 1975 O† a tangent. J Food Sci 40 410È412.Martin F W 1983 The carotenoid pigments of white Ñeshed

sweetpotatoes. J Agric Univ Puerto Rico 67 494È500.McGuire R G 1992 Reporting of objective color measure-

ments. HortSci 27 (12) 1254È1255.Ong S H, Tee E S 1992 Natural sources of carotenoids plants

and oils. In : Methods in Enzymology (Vol 213 : CaroteoidsChemistry, Quantity and Antioxidation), ed Parker L. pp142È167. Academic Press, New York.

SAS 1994 SAS UserÏs Guide (Version 6, Fourth Edition). SASInstitute Inc, Cary, NC, USA.

Takahata Y, Noda T, Nagata T 1993 HPLC determination ofb-carotene of sweetpotato cultivars and its relationship withcolor values. Jpn J Breed 43 421È427.

Woolfe J A 1992 T he Sweetpotato : A Neglected Crop. Cam-bridge University Press, Cambridge, UK.

Worthington O J 1954 Color di†erences in the quality evalu-ation of processed fruits and vegetables. A symposium oncolor in foods. Quartermaster Food and Container Institutefor the Armed Forces.